The New Era of Hybridisation and Learning in Heuristic Search Design

Abstract This chapter aims to extend on the overview of heuristic and metaheuristics described in chapter [51] by focussing on the new developments related to hybridisation and learning when designing an effective heuristic, metaheuristic, or machine learning technique. This will include a wider discussion on hybridisation, deep learning and a brief on machine learning and big data. Some of the mechanisms that enhance their implementation by turning these heuristic-based techniques into powerful, efficient and practical optimisation/statistical tools are discussed. This is attributed to the incorporation of speed-up mechanisms that can be inspired from data structure, neighbourhood reduction, cost function approximation, parallelisation among others. The chapter also provides a highlight of potential research avenues that can be worth exploring.